Research Recommendations Following the Discovery of Pain Sensitizing IgG Autoantibodies in Fibromyalgia Syndrome

Abstract Background Fibromyalgia syndrome (FMS) is the most common chronic widespread pain condition in rheumatology. Until recently, no clear pathophysiological mechanism for fibromyalgia had been established, resulting in management challenges. Recent research has indicated that serum immunoglobulin Gs (IgGs) may play a role in FMS. We undertook a research prioritisation exercise to identify the most pertinent research approaches that may lead to clinically implementable outputs. Methods Research priority setting was conducted in five phases: situation analysis; design; expert group consultation; interim recommendations; consultation and revision. A dialogue model was used, and an international multi-stakeholder expert group was invited. Clinical, patient, industry, funder, and scientific expertise was represented throughout. Recommendation-consensus was determined via a voluntary closed eSurvey. Reporting guideline for priority setting of health research were employed to support implementation and maximise impact. Results Arising from the expert group consultation (n = 29 participants), 39 interim recommendations were defined. A response rate of 81.5% was achieved in the consensus survey. Six recommendations were identified as high priority- and 15 as medium level priority. The recommendations range from aspects of fibromyalgia features that should be considered in future autoantibody research, to specific immunological investigations, suggestions for trial design in FMS, and therapeutic interventions that should be assessed in trials. Conclusions By applying the principles of strategic priority setting we directed research towards that which is implementable, thereby expediating the benefit to the FMS patient population. These recommendations are intended for patients, international professionals and grant-giving bodies concerned with research into causes and management of patients with fibromyalgia syndrome.

undertook a research prioritisation exercise to identify the most pertinent research approaches that may lead to clinically implementable outputs. Methods. Research priority setting was conducted in five phases: situation analysis; design; expert group consultation; interim recommendations; consultation and revision. A dialogue model was used, and an international multi-stakeholder expert group was invited. Clinical, patient, industry, funder, and scientific expertise was represented throughout. Recommendation-consensus was determined via a voluntary closed eSurvey. Reporting guideline for priority setting of health research were employed to support implementation and maximise impact. Results. Arising from the expert group consultation (n ¼ 29 participants), 39 interim recommendations were defined. A response rate of 81.5% was achieved in the consensus survey. Six recommendations were identified as high priority-and 15 as medium level priority. The recommendations range from aspects of fibromyalgia features that should be considered in future autoantibody research, to specific immunological investigations, suggestions for trial design in FMS, and therapeutic interventions that should be assessed in trials. Conclusions. By applying the principles of strategic priority setting we directed research towards that which is implementable, thereby expediating the benefit to the FMS patient population. These recommendations are intended for patients, international professionals and grant-giving bodies concerned with research into causes and management of patients with fibromyalgia syndrome.
Key Words: Fibromyalgia; Autoantibodies; Research Priorities Background Fibromyalgia syndrome (FMS) is a widespread chronic pain condition associated with multimodal sensory hyperresponsiveness, fatigue, and changes in cognition, which typically affects mood and/or function [1,2]. The World Health Organization (WHO) ICD-11 has classed FMS as "chronic primary pain." FMS is common, with an estimated prevalence of between 2% and 6% in the general population worldwide, depending on criteria used [3]. It imposes a major burden on affected individuals, the healthcare system, and the general economy (Silverman et al. 2009). The pathophysiology of FMS is poorly understood, and available treatments are insufficiently effective [4]. An absence of specific disease biomarkers hinders patient stratification and impacts on all areas of research into FMS.
Recently, a consortium of UK and Swedish investigators has reported research findings indicating that noninflammatory immunoglobulin G (IgG) serum autoantibodies (Aab), without a systemically measurable inflammatory response contribute to the pathogenesis of FMS [5]. The team transferred both single and pooled serum-IgG from patients with FMS to rodents and identified typical features of clinical FMS in the rodents, including sensitivity to pressure and cold, reduced grip strength and movement, and small fiber pathology when compared to the transfer of serum-IgG from healthy volunteers. The antibodies were shown to specifically bind to structures within mouse and human dorsal root ganglia. These findings may provide a first pathogenic clue to the pathophysiology of FMS and necessitate questions about how the pain research community can use this new knowledge to leverage and improve research in this area.
In order to address the implications of these findings for FMS research an expert multidisciplinary working group was convened. The aim of this group was to develop patient-centred, translatable, and clinically useful research priority recommendations based on these novel findings in FMS. Here we outline these recommendations and detail their development.
The initiative was developed with the aim of formulating research recommendations arising from the recent discovery of pronociceptive Aab in FMS [5]. We adapted the REPorting guideline for PRIority Setting of health research (REPRISE) [6] to facilitate comprehensive reporting of the development of, and conclusions arising from, the expert working group in FMS.
These recommendations are primarily intended for professionals and grant-awarding bodies internationally concerned with research into the causes and management of patients with FMS, to ultimately benefit this patient population. A revision is planned in 5 years' time.

Methods
The reporting of this study follows the REPRISE and CHERRIES guidelines [6,7].

Governance and Team
The initiative was developed by one of the authors (AG) after consultation with co-authors in the antibody publication. Additionally, stakeholders were invited that were independent FMS experts or experts in autoantibody research as identified by these authors, as well as the Funder (Versus Arthritis [VA]), the UK's largest charity dedicated to supporting people with arthritis, rheumatic diseases, and musculoskeletal disorders such as FMS, which also has extensive experience in research priority initiatives https://www.versusarthritis.org/. Senior functional leaders from Eli Lilly, a large pharmaceutical company with an established autoimmune and pain portfolio were invited to contribute industry stakeholder perspectives. Administrative support was provided through a VA administrator, and an administrator funded via the Pain Relief Foundation, Liverpool, a dedicated UK Pain Research charity https://painrelieffoundation.org.uk/ (HMC). A researcher independent from VA with experience in leading research priority setting exercises (ED) supported the output write-up.

Public and Patient Involvement
Versus Arthritis was involved throughout this project. In addition, two people living with FMS and one familial carer were members of the working group. These patient research partners were invited from participants in an existing FMS patient involvement initiative at the University of Liverpool, and from amongst a patient involvement group within VA. They took part in all discussions, feedback, consensus survey and reviewed the manuscript. Patient research partners were offered assistance with the consensus survey, including assistance with the language, terminology, and technology used.

Process Framework
We followed a dialogue model [8] with main principles including respect for the experiential knowledge of all the participating stakeholders. The focus of the dialogue model is improved stakeholder learning opportunities through dialogic processes that explore and move beyond initial assumptions about facts and values surrounding complex problems. Briefly, through the process of i) situation analysis, ii) design, iii) expert group consultation, iv) interim recommendations, and v) consultation and revision (outlined in greater detail in the next section) stakeholders learn from each other the different ways that a shared problem can be defined through reframing and reformulating stakeholder issues, problems, and questions. This emphasizes mutual learning between stakeholders of all types and enhances shared understanding of the issue at hand for all stakeholders.

Research Recommendation Process
The initiative was conducted in five phases, i) situation analysis, ii) design, iii) expert group consultation, iv) interim recommendations, and v) consultation and revision ( Figure 1). Situation analysis was conducted by the convenors defining the scope. Design phase included stakeholder engagement, protocol development, and defining decision making criteria. The expert group consultation is outlined in detail below. This consisted of a 2-day face-toface workshop of invited multidisciplinary researchers, clinicians and people living with FMS. It was facilitated by the research charity funders (VA). Interim recommendations arising from the workshop were prepared and shared with the stakeholders who had attended the meeting. Stakeholders were invited to critically assess these recommendations, which were reviewed in an iterative manner until consensus was reached. After two iterative rounds, an anonymous e-survey was conducted to ascertain levels of stakeholder agreement.

Stakeholders/Participants
The group consisted of 29 stakeholders (16 male, 13 female), of which n ¼ 25 also took part in the face-to-face workshop; it included patients (n ¼ 2) and caregivers (n ¼ 1). One of the participating patients had previously received specific research review training from VA and had experience as a "patient insight partner" for painrelated research. The group provided a broad mix of clinical and scientific skills and experiences (Table 1). Additionally, stakeholders' countries of residence had a wide geographic spread (UK 21 of which England 18, Scotland 2, Wales 1, France 1, Germany 2, Hungary 1, Ireland 1, Sweden 2, USA 1). These were 27 established researchers or consultant level clinicians and two early career researchers. These participating researchers were included based on their relevant experience, including expertise in Aab research, Aab contribution to FMS, the mechanisms underpinning rheumatological or neurological autoimmune disorders, or a track record of research activity in fibromyalgia and/or chronic pain. Three representatives from the research department of a pharmaceutical company with an active pain portfolio also took part; two staff members of the Funder's research and discovery department participated in the workshop, and additionally a VA administrator (as a nonstakeholder). All stakeholders were invited to contribute to defining the final meeting agenda. All travel expenses were paid to the nonindustry participants, as was overnight accommodation if needed, no honoraria were offered.

Situation Analysis and Consultation Design
A draft content outline of potential research topics that may need discussion was formulated by AG integrating similar topics addressed in other autoimmune conditions and modelled on previous focus groups on immune trials in chronic pain conditions and multi-stakeholder processes in pain conditions [9,10]. This outline and a preliminary list of stakeholders was circulated among two of these stakeholders (D.A., C.B.) who made suggestions about content and involvement of additional stakeholders; a revised content proposal was sent to the identified stakeholders and changes (both regarding content and invitation of additional stakeholders) were implemented and a final version circulated amongst all stakeholders. An agenda for a two-day workshop was condensed from that document; the agenda was circulated in advance to allow a wider stakeholder review, and revisions were implemented where necessary. The finalized agenda included 11 topic areas for discussion.
The aim for the workshop was to include discussion of the full range of suggested topic areas without any ranking or exclusion. The 11 topic areas are outlined in Table 2. At the face-to-face workshop each topic area was introduced by one stakeholder with expertise in that field, and then deliberated with the working group using a roundtable discussion format.

Analysis and Identification of Research Recommendations
Minutes from the working group meeting were taken independently by three people. These minutes were then combined. In preparing the overall meeting minutes, the extensiveness, intensity, and specificity of comments at the meeting were considered. Draft minutes were circulated to the entire working group for review to minimise bias and ensure they accurately reflected the working group discussion. The minutes were used to identify research recommendations across all 11 topics discussed. Recommendations were identified. These recommendations were grouped into seven broad areas (Table 3). Feedback and consensus on these 39 recommendations were assessed via an online survey.

Consensus eSurvey
Quota sampling of the working group was used for a voluntary closed survey hosted on the survey monkey platform. The survey consisted of 40 questions. The survey was distributed across nine pages. Page 1 was the introduction; pages 2-8 were the 39 recommendations (one category per page), and page 9 was a thank you and a "further comments" question. Participants could review and change their answers. The research recommendations were assessed via a 5-point Likert scale (Strongly Agree; Agree; Neither Agree nor Disagree; Disagree, and Strongly Disagree).
The eSurvey was reviewed by one native and one nonnative English speaker. Participants were informed of the intended purpose of the survey and that it was anonymous. The survey was open for 20 days, starting on 23 April 2020. It was advertised via email to the working group members. Three emails were sent, each four days apart, to encourage participation. The survey link was included in each email. No incentives were offered for completion. Unique survey respondents were measured via IP address. Cookies were not used.

Survey Analysis
An IP check was performed to identify potential duplicate entries. No duplicate entries were identified. All questionnaires, including incomplete questionnaires, were analyzed. "Strongly agree" and "agree" responses were combined as "agreement," and all other responses were considered as disagreement. Data were coded and SPSS (v24.0) was used to preform frequency analysis. Agreement data are reported as count (n), percentage agreement of respondents, and percentage agreement of total cohort.

Results
Two members of the working group self-selected out, to avoid potential perceived bias (one funder and one pharmaceutical industry representative). The cohort was n ¼ 27. Of these, n ¼ 22 responded, giving a response rate of 81.5%. As 81.5% was the highest achievable percentage of agreement for the total cohort, the consensus for each recommendation was categorised based on percentage agreement of total cohort: high consensus (>72%), medium (62%-71%), low (51-61%), and less than 51% was deemed to have no consensus. Six recommendations achieved high consensus, whereas n ¼ 15 had medium level of consensus (Table 4). Research recommendations with high or medium stakeholder agreement fell into the following five thematic groups:

Facets of FMS
This category includes recommendations about aspects of fibromyalgia features and associated symptoms that should be considered in all future Aab research.  [5]. 5. The contribution of Aab to chronic pain conditions associated with "fibromyalgianess," below the ACR 2016 score threshold for the diagnosis of FMS [12], and in chronic overlapping pain conditions (COPCs) should be investigated. Fibromyalgianess is an emerging concept; it describes a symptom complex variably comprised of fatigue, sleep disturbance, cognitive symptoms, polysensory hyper-responsiveness, and psychological co-morbidities. Risk factors for this phenotype include female sex, regional or widespread pains and exposure to stressors, but it seems independent of both known peripheral triggers or individual psychopathology [13]. COPCs include conditions such as chronic pelvic pain, painful bladder syndrome, irritable bowel syndrome, low back pain, burning mouth syndrome, post-traumatic chronic pain, chronic fatigue, and others [14]. It is more rule than exception that people with FMS have suffered or are suffering from COPCs, and how antibodies contribute to these pain phenotypes (both with and without presence of FMS) would be important to investigate.

Aab Research
There was a high level of stakeholder agreement that the recent Aab findings should be repeated by independent researchers and research laboratories. It was acknowledged that the current findings already incorporate some degree of replication from two independent laboratories and two distinct patient cohorts. Stakeholders considered characterisation of Aab targets as a key relevant objective for Aab research going forward, with a high consensus level. There was also a good level of agreement that for each identified Aab target the antibody pathogenicity should be investigated; it was noted that specific Aab associated with a disease can sometimes be without a clear pathogenetic role.
Stakeholders considered it important to research if and how Aab can cause central phenomena that characterize FMS, and to further investigate Aab binding in the central nervous system (CNS). Stakeholders acknowledge Preclinical research into Aab mechanisms in FMS.

3.
The role of B-cells in FMS.

5.
Is there a relationship between small fibre neuropathy and pronociceptive Aab in FMS? 6.
Assessment of functional brain activation and research into brain-targeting interventions. 7.
Potential new or adapted clinical treatments for FMS in light of Aab findings. 9.
Trial design for FMS trials. 10.
Research recommendations to investigators in independent trials in other conditions, to capture prevalence and response of comorbid FMS. 11.
How patient involvement can enhance research.
Aab ¼ autoantibody; FMS ¼ fibromyalgia syndrome.   that no evidence for CNS binding has been identified but advised that this may not exclude binding at lower intensities or to distinct, small regions.

Immunology
There was strong agreement that immune parameters should be assessed during FMS flares versus non-flare periods; the dominant antibody subclasses and their titres should be studied [15]. It was thought possible that key parameters such as antibody levels or plasmablast populations may differ during flare times. This is also a method to facilitate generation of antibodies in vitro, from patient B cells. Stakeholders also agreed that the mechanisms underpinning the production of noninflammatory Aab should be investigated.
In clinical trials, Aab serum levels-alongside the above immunological parameters-should be ascertained both at baseline and post-intervention; this includes both pharmaceutical, and non-pharmaceutical intervention (e.g., behavioural) trials.

Trial Design in FMS Trials
Although not specific to Aab research, stakeholders considered the design of clinical trials; one consideration was that side effect profiles for any immunological drug trials addressing Aab mechanisms would differ from currently used drugs, and trial designs would need to address this.
There was high consensus that primary endpoints for interventional trials in FMS should reflect patients' multi-dimensional experiences, which often include symptoms such as cold sensitivity, pressure sensitivity, emotional distress, and fatigue, rather than solely focusing on the perceived pain intensity [16]. Stakeholders discussed how their consensus is best reconciled with traditional recommendations of using pain intensity as the primary outcome for pain-condition trials. A discussion within the worldwide research community was suggested important.
It was agreed that FMS interventional trial designs should generally be co-created both with patients and interdisciplinary professionals.
Stakeholders agreed that research into decision making by people with FMS about their participation in interventional trials would be important, to employ methods that truly empower patients in their decision making, particularly in the context of drugs which can have serious adverse effects [9,17].
"Placebo-dread" may affect trial recruitment and outcomes in FMS trials, where patients fear that reporting a good improvement of their symptoms may later be interpreted as being "all in their head" if randomized to placebo; this may be particularly important in investigatorled trials where patients may know the principal investigator from clinical interactions. Such placebo-dread might lead to under-reporting of good outcomes. Little research exists in this area, and more is needed. Stakeholders agreed that this issue should be addressed in FMS trial designs. Good patient preparation before trial entry should go some way to minimise this issue; for example, conveyance of an understanding of the universal nature of the placebo effect (not just limited to FMS patients) may be useful. Independently, it was also noted that longer FMS durations may be associated with lower placebo responses [18].

Suggestions for Clinical Trials in FMS
This topic was discussed extensively, and good agreement was found for only one recommendation. Stakeholders agreed that a three-arm experimental trial including immunoglobulin G (IgG) immunoadsorption versus plasma exchange should be considered to improve understanding of the clinical relevance of pain-sensitizing Aab in FMS. Patients would need to be told before enrolment that even if that trial was successful these interventions may not become available in clinical practice. Plasma exchange (removal of serum and all solutes from the blood) and immunoadsorption (removal of IgG or IgG subclasses through specific columns) are often effective in Aab-mediated autoimmune conditions [19]. However, these interventions are cumbersome given they require venous access and are generally considered less suited for long-term treatment. They are frequently being used to interrupt disease flares or to rapidly bring an immune disorder under control until the effect of disease modifying treatments set in. There was agreement that FMS patients enrolled in interventional trials aiming to address Aab or their downstream mechanisms should be tested for the presence of such antibodies; serum tests are yet to be developed-passive transfer bioassay testing of each screened patient is unlikely feasible or ethically acceptable.

Small Fiber Neuropathy
Recently, several research groups have established that small fiber pathology often affect patients with FMS, with a correlation to the FMS disease severity in smaller subgroups [20]. At the same time, nociceptor excitability measured using microneurography is often abnormal [21]. Correspondingly, investigations in the recent FMS rodent-transfer model included assessment of small fibre neuropathy following Aab transfer, with results indicating that dermal nerve fibre density is indeed reduced in the mouse hind paw skin from at least as early as 4 days after transfer start, although behavioural abnormalities are present from at least 1 day after administration of human IgG. This raises the question of whether the abnormal rodent behaviour is secondary to the developing structural nerve fiber abnormalities, or alternatively whether it is due to primarily functional changes induced by the Aab. The latter principle has recently been demonstrated in a different, very rare Aab mediated, painful neurological condition [22]. Stakeholders agreed that the hypothesis that Aab can induce abnormal nerve fibre activation independent of any structural nerve fibre change should be further investigated, and also that the relationship between small fibre neuropathy and the presence of pain-sensitizing Aab should be addressed.

Additional Points of Discussion
Some discussion points not included in the survey were thought relevant by the organising team; in part they relate to matters of fact, or to notes made during the workshop. We outline these here, highlighting that the level of agreement was not assessed: • The degree of fibromyalgianess is directly associated with unfavourable health outcomes, independent of the diagnosis of FMS [23][24][25][26]. • Patients suffering from a higher number of COPCs may have higher fibromyalgianess scores, but this has not been assessed systematically. • Patients with high symptom severity scores (SSS) may experience more stable FMS than others. • Clinical assessment of comorbid FMS can support rational treatment; for example, in rheumatoid arthritis may avoid successive biological treatments to address patient pain complaints caused by FMS [27,28]. • Possible immune contribution in FMS has previously been suggested indirectly by studies assessing the efficacy of IVIG treatment in a subgroup of non-comorbid FMS patients [29,30], in FMS patients co-morbid with polyneuropathy [31], and by demonstration of abnormal IgG skin deposits in FMS [32]. • The transfer model cannot easily model long-term changes, due to the development of serum sickness in the animals arising from prolonged human protein transfer. • Antibody profiles, specificities and/or functional effects in FMS differ from those earlier described in other chronic pains conditions, for example Complex Regional Pain Syndrome (CRPS) [33] and non-inflammatory joint pains in RA [34], suggesting that specific Aab may cause specific chronic pain phenotypes. • Clinical methods to better understand the role of small fibre pathology may include longitudinal investigations, for example, in the context of effectiveness trials (repeat skin biopsies, or corneal microscopy at baseline/follow-up). More information about the relation between small fibre changes and fatigue would also be important. • Anecdotal evidence from patients with co-morbid FMS treated for their rheumatological condition with biologics including anti-cytokine agents, or B-cell therapies suggest that FMS condition may not readily respond to these therapies; independently, the situation in axial spondylarthritis may be more complex, as in this condition fulfilling FMS criteria may not necessarily mean that FMS is truly present, and the symptom severity score may be a better measure [35]. More systematic understanding of these issues is required to avoid potentially futile trial efforts. • New, effective immune interventions in FMS patients, unless achieving complete remission, may challenge trial-patients' established activity-pacing skills. Patients who have less pain on drug may overdo their activities resulting in pain increase, potentially causing bias in clinical trials. Clinical trials should consider including multidisciplinary support for pacing daily activities at the respective new pain thresholds.
• Multidisciplinary preparation prior to a trial may also allow patients to put symptoms experienced during the trial, both beneficial and adverse, into context, resulting in better adherence and a less stressful experience [17]. • Inclusion of PainDetect [36] and "First" [37] to the outcome questionnaires in FMS trials may provide further relevant information. • Symptoms and diseases can be understood within an evolutionary framework as arising from adaptive defences [38]; this includes depression [39] and the framework could usefully be applied to the auto-antibody production in FMS

Limitations
One limitation was that we included relatively fewer early career researchers or trainee clinicians when compared to established researchers. Future processes might enhance inclusiveness and multifaceted responses by including additional participants from the former group.

Conclusion
Recent findings of pain-sensitising Aab in FMS open a new field of research, with questions extending across a wide range of areas. In this focussed research priority setting exercise, stakeholders from a mix of professional backgrounds and world-regions, funders, relatives, and patients with FMS agreed on a set of research recommendations. These proposed research areas might be taken forward by interested researchers and funders worldwide. Stakeholders agreed that the confirmation of Aab contributing to the cause of FMS would constitute a transformative finding for this field. It would also represent a revolutionary advance that should remove the perceived stigma arising from the current lack of understanding often communicated by patients [40].
Ensuring that research is centred on improving FMS management and is translatable to a clinical context will be critical to improve the quality of life for this historically undermanaged patient cohort [41].

Ethics Approval and Consent to Participate
Not applicable. Consent for publication. Not applicable.
Availability of data and materials. The data sets generated and/or analyzed during the current study are available in the Figshare repository, [DOI: 10.6084/m9.figshare.12602186 (will become activated following acceptance for publication)]